Pull Request已成功合入, 合并人@ascend-robot
(感谢 Ambi 的贡献)Thanks for your pull-request.
The full list of commands accepted by me can be found at here。
You can get sig-info at here
PR Approval Progress
✅ Congratulations! All modules have met the lgtm and approve requirements.
Module Approval Details
| module | lgtm status | approve status |
|---|---|---|
| test | ✅ zichun_ye, adrian07110 (2/2) | ✅ zichun_ye, adrian07110 (2/1) |
| torch_npu/utils | ✅ zichun_ye, adrian07110 (2/2) | ✅ zichun_ye, adrian07110 (2/1) |
💡 Tip:
- Committer can comment
/approveor/lgtm- Commenting
/approveimplies both code review (lgtm) and intent to merge (approve)
CLA Signature Pass
sunhaolun, thanks for your pull request. All authors of the commits have signed the CLA. 👍


当前仓库存在以下 保护分支 :
| Protected Branch | Version | Release |
|---|---|---|
| master | ||
| v2.10.0 | ||
| v2.7.1 | ||
| v2.9.0 | ||
| v2.11.0 | ||
| v2.12.0 |
评论 /sync <branch1> <branch2> ... 可将当前 PR 修改同步到其它分支(创建同步 PR):
a) 如果当前 PR 是 Open 状态,同步操作将延迟到 PR 被合并时执行
b) 如果当前 PR 已经 Merged,将立即执行同步操作
注意:
- /sync 命令可以指定同步到多个分支,仅最后一个 /sync 命令生效
- 如果创建的同步 PR 不正确,可通过向同步 PR 的源分支提交轻量级 PR 完善,或使用 /close 命令关闭


compile


通过compile命令启动流水线PR-pipeline_pytorch失败:流水线并发数超限,请稍后重试


Ascend docs pipeline is running...


✅ 跳过 docs ci 检查,没有需要检查的文档文件


Ascend docs pipeline is running...


compile


Ascend docs pipeline is running...


✅ 跳过 docs ci 检查,没有需要检查的文档文件


| 阶段 | 任务名 | 状态 | 详情 |
|---|---|---|---|
| 编译构建 | Build_X86 | ✅ | >>> |
| Build_ARM | ✅ | >>> | |
| Build_LibTorch_x86 | ✅ | >>> | |
| Build_LibTorch_ARM | ✅ | >>> | |
| Build_X86_torchair | 🛑 | >>> | |
| Build_ARM_torchair | 🛑 | >>> | |
| patch_test | 🛑 | >>> | |
| 恶意代码检查 | Antipoison | ✅ | >>> |
| 编码安全与规范检查 | CodeCheck | ✅ | >>> |
| check_error | ✅ | >>> | |
| CodeCheck_lintrunner | ✅ | >>> | |
| 开源片段检查 | SCA | ✅ | >>> |
| 开发者测试 | UT_X86_Part_01 | 🛑 | >>> |
| UT_X86_Part_02 | 🛑 | >>> | |
| UT_ARM_A3_Part_01 | 🛑 | >>> | |
| UT_ARM_A3_Part_02 | 🛑 | >>> | |
| UT_ARM_A2_Part_01 | ✅ | >>> | |
| UT_ARM_A2_Part_02 | ✅ | >>> | |
| UT_ARM_A2_Part_03 | ✅ | >>> | |
| UT_inductor_Part_01 | 🛑 | >>> | |
| UT_inductor_Part_02 | 🛑 | >>> | |
| UT_inductor_Part_03 | 🛑 | >>> | |
| UT_inductor_Part_04 | 🛑 | >>> | |
| UT_DIST_ARM_Part_01 | 🛑 | >>> | |
| UT_DIST_ARM_Part_02 | 🛑 | >>> | |
| UT_DIST_ARM_Part_03 | 🛑 | >>> | |
| UT_DIST_ARM_Part_04 | 🛑 | >>> | |
| 流水线 | PR-pipeline_pytorch | ✅ | >>> |
- compile : 运行流水线
- retry : 重试流水线所有失败子任务
- retry <任务名> : 仅重试指定失败子任务
- stop : 停止流水线


/approve


/approve


Review Guide
This pull-request passes review.
Committers who wrote a comment of /approve are: zichun_ye, adrian07110.
Reviewers who wrote a comment of /lgtm are: adrian07110, zichun_ye.




https://gitcode.com/Ascend/pytorch/issues/1962
【合入来源】
【修改方案】
torch_npu/utils/_dynamo.py 的 patch_inductor_wrapper() 将 new_get_config_copy 通过 类级别 (ConfigModule.get_config_copy = ...) patch 到了所有 ConfigModule 实例上。当 torch._functorch.config.get_config_copy() 在第一次编译过程中被调用时(jit_compile_runtime_wrappers.py:194),4 个 NPU 特有的 config key 被永久注入到 functorch config 的 _config 字典中。第二次编译时 save_config() 序列化的内容多了这 4 个 key → 不同的 hash → AOT autograd cache miss。
修复:在 new_get_config_copy 中添加一行检查 if self is not torch._inductor.config: return ori_dict,确保 NPU config key 只注入到 inductor config,不影响 functorch config 等其他 ConfigModule 实例。
【资料变更】
不涉及
【接口变更】
不涉及
【功能验证】
全量UT验证
【CheckList】